[DocDB] QLTabletTest.GCLogWithoutWrites is flaky in fast debug builds

[rthallamko3]

Jira Link: DB-5921

Description

Detective report

Failure trends

It seems like the validation of log operations after leader step down needs to be adjusted for fastdebug builds at least.

TEST_F(QLTabletTest, GCLogWithoutWrites) {
  SetAtomicFlag(kRetryableRequestTimeoutSecs, &FLAGS_retryable_request_timeout_secs);

  TableHandle table;
  CreateTable(kTable1Name, &table);

  FillTable(0, kTotalKeys, table);

  std::this_thread::sleep_for(1s * (kRetryableRequestTimeoutSecs + 1));
  ASSERT_OK(cluster_->FlushTablets());
  DoStepDowns(cluster_.get());
  VerifyLogIndicies(cluster_.get());
}

void VerifyLogIndicies(MiniCluster* cluster) {
  for (size_t i = 0; i != cluster->num_tablet_servers(); ++i) {
    auto peers = cluster->mini_tablet_server(i)->server()->tablet_manager()->GetTabletPeers();

    for (const auto& peer : peers) {
      int64_t index = ASSERT_RESULT(peer->GetEarliestNeededLogIndex());
      ASSERT_EQ(peer->consensus()->GetLastCommittedOpId().index, index);
    }
  }
}

Warning: Please confirm that this issue does not contain any sensitive information

• I confirm this issue does not contain any sensitive information.